Mobile Termination and Mobile. Penetration

Transcription

1 International Conference on Infrastructure Economics and Development, 2010 (Toulouse) Sjaak Hurkens (IAE, (CSIC) and IESE (SP-SP)) Doh-Shin Jeon (TSE, CEPR) 1

2 Network competition: (two-way) access pricing Retail price ORIGINATING OPERATOR Access price TERMINATING OPERATOR Cost of origination Cost of termination 2

3 The literature on two-way access Starters: Armstrong (1998) and Laffont-Rey- Tirole (LRT)(1998a,b) Study (i) How does a given reciprocal access price affect competition among network operators and hence the retail tariffs? (ii) What is the optimal access price? Many extensions Approach based on fixed per-minute access price 3

4 Motivation from one-way access ECPR (Efficient Component Pricing Rule) (Willig 1979, Baumol and Sidak 1994): The access charge depends on the retail price (p) of the incumbent: a - c 0 = p - c 4

5 ? Make access price depend on retail price 5

6 Retail benchmarking approach to two-way access (Jeon-Hurkens, RAND 2008) We propose a retail benchmarking approach to two-way access pricing When networks use two-part tariffs, our approach achieves the static efficiency as well as the dynamic efficiency. 6

7 What we do in this paper Set up: elastic subscription demand, on-net and off-net price discrimination, three-part tariffs Standard approach: fixed (per-minute) and reciprocal termination charge Q1: Which level of termination charge do the firms prefer? Q2: Which level of termination charge does the regulator prefer? Retail benchmarking approach Q3: Can we do better and how? 7

8 Elastic subscription demand: mobile penetration rates (source, ITU 2007) Japan: Korea: USA: 86 Africa: (mobile), 3.21 (fixed) India: (mobile), 3.93 (fixed) 8

9 Key result The retail benchmarking appoach allows to increase subscriptions without distorting marginal prices by intensifying competition in terms of fixed fees. 9

10 Outline 1. Logit Model with Rational expectations 2. Fixed Access Charges 3. Retail Benchmarking Rule 4. Policy implications 10

11 Model: cost Costs: c the marginal cost of a call c T (<c) the marginal cost of terminating a call On-net marginal cost: c Off-net marginal cost: c- c T +a 11

12 On-net vs off-net perceived marginal cost c c-c T +a 12

13 Model: Demand and tarifs - Mass 1 of consumers - u(q): concave increasing utility from calls of length q - Duopolists set ( F, p, pˆ i i i ) 13

14 Model: Demand for subscription Given tariffs, consumers form expectations about number of subscribers Rational expectations equilibrium in a Logit model 14

15 Fixed Access Charge a : perceived marginal cost pricing Maximizing profit with respect to usage prices leads to perceived marginal cost pricing: p i = c, pˆ i = c + a c T 15

16 Business stealing vs. network externality An increase in F 1 has an ambiguous effect on the number of subscribers on 2 α α2 F1 α2 F 1 > < 0 0 Net Business Stealing Net Network Externality 16

17 Fixed Access Charge a : General result Firms always want to have acess charge below c T for different reasons Case 1 (net business stealing effect): in order to soften competition Case 2 (net network externality effect): in order to internalize better the network externality effect. 17

18 Retail Benchmarking Approach Information constraint of the regulator: the regulator - Does not know the demand structure and the fixed cost - Knows the marginal cost structure - Observes the retail tariffs (as consumers do) 18

19 Retail Benchmarking Approach Retail profit per customer π i ( a) Define access charge to be paid by network i to a rival network; π i( a) ai = ct + k q ( p ˆ ) i 19

20 Retail Benchmarking Approach We still have perceived marginal cost pricing. Furthermore, Π i = αi[ Fi + α jr( cˆ) f ] kαiα j[ Fi Fj ] Competition in fixed fee, fiercer for larger k>0! 20

21 Retail Benchmarking Approach Since fixed fee decreases in k, we can increase penetration without (further) distorting usage prices 21

22 Conclusion Retail benchmarking approach works better than the standard approach: k(>0) increases competition in fixed fees and thus increases penetration without distorting usage prices 22

23 Policy Implications Australian Competition and Consumer Commission (2001) adopted a retail benchmarking approach But their rule was like the ECPR After the introduction of the approach, some mobile companies raised their retail prices and the commission abandoned the approach Linking access charges to retail prices in a right way intensifies retail competition 23